| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the VA Connecticut Healthcare System (M.M.B., M.C.B., R.R., R.S.), West Haven Campus, West Haven; and the Section of Cardiovascular Medicine (M.M.B., R.S.), Yale University School of Medicine, New Haven, Connecticut.
Address reprint requests to: Matthew M. Burg, PhD, VA Connecticut Healthcare System/116B4, 950 Campbell Ave., West Haven, CT 06516. Email: mburg{at}attglobal.net
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
METHODS: Medical, surgical, and psychosocial risk factors were assessed before surgery in 89 male veterans undergoing CABG. In addition, patients completed the Beck Depression Inventory. Medical, surgical, and psychological/quality-of-life outcomes were determined at 6 months of follow-up by telephone interview and review of medical records.
RESULTS: Of the 89 patients studied, 25 scored 10 or greater on the Beck Depression Inventory. The study population was dichotomized on the basis of this cutoff point. A medical prediction model was developed for each outcome of interest, based on the range of medical, surgical, and psychosocial risk indices assessed. The dichotomized depression index was added to these prediction models as a final step. This depression index was found to independently predict cardiac hospitalizations at 6 months (
2 = 4.24, p < .04), continued surgical pain at 6 months (2 = 6.36, p < .01), and failure to return to previous activity at 6 months (2 = 15.04, p < .0001). Presurgical depression also predicted depressed affect at 6 months (2 = 13.16, p < .0003).
CONCLUSIONS: Depression is an important independent contributor to medical and psychosocial morbidity up to 6 months after CABG. These findings warrant replication with larger and more diverse populations.
Key Words: depression • coronary artery bypass graft surgery • prognosis.
Abbreviations: BDI = Beck Depression Inventory; CABG = coronary artery bypass graft surgery; CAD = coronary artery disease; CES-D = Center for Epidemiological Studies Depression Scale; CHF = congestive heart failure; MI = myocardial infarction.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Coronary artery bypass graft surgery (CABG) is the treatment of choice for many patients with severe coronary artery disease (21), and more than 600,000 operations are now performed annually in the United States (22). The primary indications for this procedure are relief of angina and improvement in quality of life. CABG clearly benefits most patients with coronary disease (23); however, lack of improvement, both medical and psychosocial, has been noted in up to 15% of patients (24), and reduced quality of life has been noted in many more (25). To more completely understand why CABG does not produce the expected results for many patients, researchers have turned their attention to depression.
Depression has been found preoperatively in up to 47% of CABG patients (26–29) and postoperatively in as many as 61% of patients (28, 29). Furthermore, if depression is present in the preoperative period, it is likely to be seen during the postoperative period as well (28–30). For example, Timberlake et al. (30) found 37% of patients undergoing nonemergent CABG to have elevated symptoms of depression on the Beck Depression Inventory (BDI) immediately before surgery. Follow-up conducted 8 weeks and 12 months after CABG revealed that 44% and 42%, respectively, of those with elevated BDI scores at baseline continued to show significant elevations, compared with only 10% and 12% of patients without elevated BDI scores at baseline. Similarly, Pirraglia et al. (31) found that 43% of patients undergoing CABG had significantly elevated scores on the Center for Epidemiological Studies Depression Scale (CES-D) before surgery, with 41% experiencing significant mood disturbance 6 months later, compared with only 9.7% of those without baseline elevations. Hence, depression is highly prevalent in patients undergoing CABG, and it remains so for up to 12 months after the surgical procedure.
Many studies have found quality of life and psychosocial functioning to be impaired after CABG in patients who evidence depression or other types of emotional distress before the surgical procedure (32–34). For example, Jenkins et al. (32) found that patients with elevated symptoms of depression immediately after CABG were more likely than patients without depressive symptoms to report continued symptoms of angina 6 months later. Similarly, Pick et al. (34) found that those with depression are more likely to report continued fatigue. But there have been few studies concerning the relationship of depression to post-CABG medical morbidity and mortality. Early studies by Kimball (35) and by Tufo et al. (36) revealed that patients with depressive symptoms before open heart surgery (for valvular disease) had higher perioperative and postoperative mortality rates; however, the type of surgery done and the substantial improvements in surgical procedure that have been developed during the intervening years make interpretation of these findings difficult today. More recently, Connerney et al. (37) assessed depression in CABG patients using a diagnostic interview. They found an almost three-fold greater risk of a significant cardiac event for patients with major depression after surgery (and before discharge) than for those without depression. The purpose of the study reported here was to expand on these findings by examining whether clinically significant preoperative levels of depression, indexed by the BDI, were associated with 6-month medical prognosis and psychosocial functioning.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Measures
A range of medical and psychosocial measures was collected as potential determinants of prognostic outcomes after CABG (Table 1). In addition, a number of short- and long-term surgical outcomes of interest were assessed (Table 4).
|
|
Surgical.
Surgical measures included urgency of surgery, number of bypass grafts, and length of pump time and surgical procedure (in minutes). Measures of perioperative and postoperative complications included arrhythmia, CHF, MI, and infection.
Psychosocial.
Demographic information was collected, including age, marital status, education, and employment status. Depression was assessed using the BDI (38), a questionnaire previously used with cardiac populations (39) and in the study of CABG patients (30). The BDI contains 21 items, each evaluating a symptom of depression, such as sad mood, pessimistic outlook, feelings of guilt, and loss of appetite. Each item contains four sentences indicating degree of severity for that particular symptom. Subjects respond by endorsing one sentence from the set of four. Cronbach’s 
for this sample was 0.87. Prior history of and treatment for depression or anxiety disorders were not assessed.
Follow-up.
A telephone interview was used to assess medical and psychosocial outcomes subsequent to surgery. Patients were contacted by phone 4 weeks and 6 months after discharge from the hospital. During these contacts patients were questioned in detail about medical and psychosocial outcomes over the intervening period. Questioning focused on whether hospitalizations had occurred and, if so, the causes of the hospitalization (eg, MI, angina, or CHF); whether the patient was experiencing surgery-related pain; whether they had returned to a previous level of functioning or physical activity; whether they were experiencing ongoing depressed mood; and whether they had been receiving any treatment for mood difficulties (eg, psychotherapy and/or pharmacotherapy). The medical record was reviewed to support the follow-up interview. Results from only the 6-month follow-up are reported here because there were too few medical end points at the 4-week follow-up to provide for meaningful analyses. In addition, with regard to hospitalizations during the follow-up period, only cardiac-related hospitalizations (eg, new MI or unstable angina) are reported here.
Procedure
Patients were approached about participation in the study at the time their surgery was scheduled and asked to provide informed consent. After providing consent they were interviewed for demographic information, medical history, and risk factor status, and given the BDI to complete. In all cases, the interview and questionnaire assessment were completed within 1 week before surgery. Subsequent to surgery, hospital charts were reviewed to obtain preoperative, perioperative, and postoperative medical and surgical information. Patients were phoned 4 weeks and 6 months after discharge for the follow-up interview. Reports of hospitalization were verified by review of the medical record, and only verified reports were used in data analyses. All collected data were entered in a computer, coded according to patient identification number. Hence, the research and treatment teams were blinded to patients’ responses and score on the BDI.
Data Reduction and Statistical Analysis
All analyses were performed using the SAS statistical package for personal computers, version 6.12. Where dependent variables were dichotomous or ordinal in nature, logistic regression was used. This method fits the regression model to an ordinal dependent variable, allowing estimation of parameters of interest (eg, depression) while controlling for other variables (eg, medical risk score). The procedure provides an estimate of each effect adjusted for all other terms in the model, testing each item by a 2 statistic, which reflects the independent contribution of that variable. Where the dependent variable was continuous, general linear regression was used.
As a first step in analyses, a medical predictor model was developed for each outcome of interest (6-month medical morbidity, return to activity, pain, and mood; days to stepdown transfer, self-administered pulmonary toilet, and hospital discharge), using the stepwise procedure. In this procedure, variables are added one by one to the model, and the F statistic for a variable to be added must be significant at the preselected level (for these analyses, p < .05). After a variable is added, the stepwise procedure looks at all the variables already included in the model and deletes any variable that does not produce an F statistic that remains significant (40). For these analyses all variables were entered together, providing for the selection of significant indices by the stepwise procedure. Variables included in these stepwise analyses as potential predictors included all the previously mentioned indices of cardiac history, cardiac status, cardiac risk factors, medical history, and surgical procedure and complications (see Table 1 for complete list). In addition, for these analyses a weighting of age, urgency of surgery, and ejection fraction were used, following the model developed by Tu et al. (41) and Weightman et al. (42) for the study of medical outcomes after CABG.
The BDI was scored, and the study population was dichotomized on the basis of a cutoff point of 10 or greater on this measure. This cutoff was selected as indicative of clinically significant symptoms of depression, consistent with the literature on depression in patients with coronary disease (7, 39). The dichotomized depression index was then entered into the previously developed medical models, forcing in those medical and surgical indices already found to be significant as predictors of the outcomes of interest.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
2 (for class variables such as medication status). No significant differences were found.
Six-Month Outcomes
At the 6-month follow-up, six patients who had scored 10 or greater on the BDI had been admitted to the hospital for cardiac reasons (two for new MI and four for unstable angina), whereas only two who had scored below 10 on the BDI had been admitted to the hospital for cardiac reasons (unstable angina). In addition, 11 patients reported continued surgical pain (7 with a BDI score >10), and 52 had returned to their previous level of physical activity (6 with a BDI score >10). Table 2 provides the breakdown of these outcomes as a function of pre-CABG depression as well as the univariate 2 and associated p values. Logistic regression used to determine medical and surgical predictors of these outcomes revealed previous MI and chronic renal insufficiency to be predictors of hospitalization for cardiac reasons; chronic tobacco use to be a predictor of continued surgical pain at 6 months; and post-operative arrhythmia and sedentary lifestyle as predictors of return to previous activity at 6 months. Adding the dichotomized depression index to these equations revealed a significant effect for depression on cardiac hospitalization at 6 months (2 = 4.21, p < .04), for continued surgical pain at 6 months (2 = 6.36, p < .01), and for failure to return to previous activity at 6 months (2 = 15.04, p < .0001). In all cases, those with depression evidenced a poorer outcome (Table 3).
|
|
2 = 13.16, p < .0003). Hence, those who evidenced significant symptoms of depression preoperatively were highly likely to report depressive affect at the 6-month follow-up (Table 3).
Surgical Landmarks
Mean pump time was 120 minutes (±35 minutes), mean length of surgery was 335 minutes (±80 minutes), and mean number of perioperative and postoperative complications was 1.2 (±1.1). Concerning postoperative landmarks, mean days to self-administered pulmonary toilet was 2.46 (±2.76), mean days to transfer to stepdown was 4.15 (±4.42), and mean days from surgery to discharge was 11.44 (±9.10). Table 4 provides the breakdown of these outcomes as a function of pre-CABG depression. Logistic regression, used to determine medical and surgical predictors of postoperative landmarks, revealed a significant effect for postoperative CHF as a predictor of days to self-administered pulmonary toilet; a significant effect for history of arrhythmia, postoperative arrhythmia, length of surgery, and days to self-administered pulmonary toilet as predictors of days to transfer to stepdown; and a significant effect for urgency of surgery and postoperative infection as predictors of length of stay after the surgery. Adding the dichotomized depression index to these models revealed an additional significant effect for depression index on days to self-administered pulmonary toilet (Table 5).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
A second finding of this study was the impact of presurgical depression on quality-of-life indices up to 6 months after discharge from the hospital. Those scoring in the depressed range on the BDI were significantly more likely to report continued surgical pain and significantly less likely to have returned to a previous level of functioning over this time frame, findings that are consistent with other reports in the literature. For example, studies have found depression before CABG to predict such quality-of-life factors as self-assessed functional status (33); illness symptom score, comprising chest pain or pressure, shortness of breath, ease of fatigue, sequelae of surgery, and sleep problems (32); and return to work and other such activities (46, 47). One of the primary goals of CABG is symptom relief with a particular focus on alleviating symptoms in the interest of enhancing quality of life. Hence, depression seems to be a significant impediment to these outcomes.
A high proportion of patients who scored positive for depression immediately before surgery were also more likely to report continued depressive affect at the 6-month follow-up. Common thinking is that although depression is prevalent in patients during acute cardiac events, the depressive symptoms quickly dissipate after the event resolves. These data would argue to the contrary. Furthermore, the depression symptoms in these patients went largely unrecognized and untreated. Given that depression had a significant impact on medical morbidity in addition to quality of life, it would seem important for treaters to identify patients with depressive symptoms before CABG so that these symptoms can be tracked and, if they persist, treated in an appropriate and targeted manner.
These findings on the impact of depression on medical morbidity and quality of life for patients after CABG are important, but several questions remain. For example, it was beyond the scope of this study to ascertain whether the 6-month cardiac hospitalizations were due to progression of coronary disease in native vessels previously found to be diseased but not revascularized, to development of new significant disease in native vessels where disease had not previously been noted, to new occlusion in the grafts placed during CABG, or to some other process associated with functioning of the heart and not necessarily the process of development or progression of coronary artery disease (CAD). This is an important question. For example, if hospitalizations were due to progression of disease in native vessels, then the relationship of depression to medical outcomes seen in the current study can be viewed within the continuum of research that has demonstrated a role for depression as a contributor to CAD progression (3). But if the hospitalizations were due to occlusions in grafts placed during CABG or to some processes other than those associated with CAD development and progression, then depression can be seen as a significant impediment to the success of CABG itself, acting through some pathway to damage bypass grafts rapidly or impair cardiac functioning. Ascertainment of which process is at work should be addressed in future studies and can be accomplished through the use of either noninvasive diagnostic methods (eg, myocardial perfusion studies or Holter monitoring) or interventional methods (eg, cardiac catheterization).
In the current study, a questionnaire (the BDI) rather than a diagnostic interview was used to assess the presence of depressive symptoms, and many of the items on this questionnaire assess somatic complaints that are readily found in patients with coronary disease in the absence of depression. Hence, the diagnostic significance of BDI score in this population was not directly assessed. The BDI, however, is a widely used instrument in both research and clinical settings for the assessment of depression in patients with coronary disease. A cutoff point of 10 for delineation of clinically significant levels of depressive symptoms is both appropriate (20, 33) and somewhat conservative (30), and scoring done in this way is predictive of both medical morbidity and mortality with this population (39). In addition, in the current study medical prediction models that tested a comprehensive range of medical and surgical factors were first developed for outcomes after CABG, and the dichotomized depression index was entered into the prediction model only after this step. The degree to which scores on the BDI were serving as a "proxy" for the somatic complaints associated with poor cardiac health can therefore be assumed to have been partialed out.
Questions about the generalizability of findings from this study to other populations can also be raised. In particular, this group of veterans was exclusively male with a mean age of 66 years. Only 37% had more than 12 years of education, more than 70% were either unemployed or retired, and only 50% were married or living with a partner. In addition, they had an average of more than two major comorbid medical conditions. A review of the literature reveals that in other large cohort studies of patients undergoing CABG (48), the population of men tends to be better educated (64.4% with >12 years of education), more likely to be employed (58%), and more likely to be living with a partner (90%). In addition, they tend to have fewer comorbid medical conditions. But CABG is now being offered to patients with these previously considered risks for poor outcome. Therefore, as CABG surgery is extended to a wider population that is similar to the group studied here (eg, those with a greater risk for poor outcome), the importance of assessing depression before surgery may be enhanced.
These caveats highlight the importance of replication with larger, more diverse populations over a longer time frame than the 6 months of the current study. In addition, the inclusion of diagnostic measures of depression and a range of physiological indices such as an electrocardiogram (to capture new, unrecognized "silent" cardiac events) would strengthen findings and help to ascertain more completely the importance of depression for medical outcomes in patients undergoing CABG. Replication of findings would then provide the basis for large-scale clinical trials. These trials could help identify the best potential methods for identifying those at risk in a cost-effective manner and the best avenues and methods for intervention.
Identification of mechanisms by which depression affects medical outcome after CABG is also important. A range of mechanisms has been described to account for the link between depression and medical outcome for patients with coronary disease. These include poorer self-care, failure to adhere to behavioral risk reduction recommendations and medical therapy, and heightened sympathetic tone, as particularly evidenced by heart rate variability and circulating levels of stress hormones (49). Recent explorations have also focused on inflammatory processes, an approach that seems particularly promising given the central role of inflammation in the development and progression of coronary disease (50) and the observation of high levels of circulating proinflammatory proteins found in patients with depression (51–53). The identification of mechanisms of action would enhance efforts to determine the best potential points and methods of intervention and provide intermediate markers of effect.
In summary, depression before CABG surgery was found to be a predictor of 6-month cardiac hospitalization and poorer quality of life up to six months after discharge. Furthermore, depression before surgery was likely to persist. Replication of these findings with a larger group over a longer follow-up period and with a focus on potential mechanisms of action by which depression can act on medical end points is warranted. These efforts will expand our understanding of the impact that depression has on prognostically important factors and guide the development of tailored therapies for those at risk.
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Received for publication July 31, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  K. E. Freedland, J. A. Skala, R. M. Carney, E. H. Rubin, P. J. Lustman, V. G. Davila-Roman, B. C. Steinmeyer, and C. W. Hogue Jr Treatment of Depression After Coronary Artery Bypass Surgery: A Randomized Controlled Trial Arch Gen Psychiatry, April 1, 2009; 66(4): 387 - 396. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. S. Bass, D. K. Attix, B. Phillips-Bute, and T. G. Monk An Efficient Screening Tool for Preoperative Depression: The Geriatric Depression Scale-Short Form Anesth. Analg., March 1, 2008; 106(3): 805 - 809. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Oxlad and T. D. Wade Longitudinal Risk Factors for Adverse Psychological Functioning Six Months after Coronary Artery Bypass Graft Surgery J Health Psychol, January 1, 2008; 13(1): 79 - 92. [Abstract] [PDF]
|
|
|
|
 |
|
 C. Rafanelli, R. Roncuzzi, and Y. Milaneschi Minor Depression as a Cardiac Risk Factor After Coronary Artery Bypass Surgery Psychosomatics, August 1, 2006; 47(4): 289 - 295. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. H. Rosenberger, P. Jokl, and J. Ickovics Psychosocial factors and surgical outcomes: an evidence-based literature review. J. Am. Acad. Ortho. Surg., July 1, 2006; 14(7): 397 - 405. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Mallik, J. A. Spertus, K. J. Reid, H. M. Krumholz, J. S. Rumsfeld, W. S. Weintraub, P. Agarwal, M. Santra, S. Bidyasar, J. H. Lichtman, et al. Depressive Symptoms After Acute Myocardial Infarction: Evidence for Highest Rates in Younger Women. Arch Intern Med, April 24, 2006; 166(8): 876 - 883. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. M. Goyal, E. L. Idler, T. J. Krause, and R. J. Contrada Quality of Life Following Cardiac Surgery: Impact of the Severity and Course of Depressive Symptoms Psychosom Med, September 1, 2005; 67(5): 759 - 765. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. M. Ho, F. A. Masoudi, J. A. Spertus, P. N. Peterson, A. L. Shroyer, M. McCarthy Jr, F. L. Grover, K. E. Hammermeister, and J. S. Rumsfeld Depression Predicts Mortality Following Cardiac Valve Surgery Ann. Thorac. Surg., April 1, 2005; 79(4): 1255 - 1259. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Mallik, H. M. Krumholz, Z. Q. Lin, S. V. Kasl, J. A. Mattera, S. A. Roumains, and V. Vaccarino Patients With Depressive Symptoms Have Lower Health Status Benefits After Coronary Artery Bypass Surgery Circulation, January 25, 2005; 111(3): 271 - 277. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. M. Burg, M. C. Benedetto, and R. Soufer Depressive Symptoms and Mortality Two Years After Coronary Artery Bypass Graft Surgery (CABG) in Men Psychosom Med, July 1, 2003; 65(4): 508 - 510. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
